Claire Vo · 2 min read · Gen AI Tools
GLM 5.2: why I’m replacing Opus in Claude Code with this new model
Read the original on lennysnewsletter.com →
This article documents a hands-on evaluation of GLM 5.2, the open-weight coding model from Z.AI, tested on four real tasks inside a production Next.js app: a codebase architecture audit, a UI redesign, and a 45-minute autonomous bug-hunting session. It covers setup in Cursor and Claude Code, performance on codebase exploration and architecture summarization, and whether GLM 5.2 can match a design system, plus a practical cost breakdown (about $3.36 for ~6 million tokens) and where it stumbled.
Key Takeaways
- Understand what 'open-weight' means and why it matters for cost and vendor independence.
- Learn how to connect GLM 5.2 to Cursor and Claude Code for coding tasks.
- See how GLM 5.2 performs on production tasks like codebase exploration, architecture summarization, and UI/page redesign, including a 45-minute autonomous run.
- Review where GLM 5.2 falls short and obtain a transparent cost breakdown for the experiment.
Excerpt
I put GLM 5.2, the open-weight coding model from Z.AI, through four real tasks inside my actual codebase: a codebase architecture audit, a UI redesign, and a 45-minute autonomous bug-hunting session pulling from Sentry and Vercel logs. Total cost: $3.36 for roughly 6 million tokens, a prioritized bug-fix dashboard I’m actually shipping from, and a landing page redesign that matched Chat PRD’s design system on the first try. Listen or watch on YouTube , Spotify , or Apple Podcasts What you’ll learn: What “open-weight” actually means and why it matters for cost and vendor independence How to connect GLM 5.2…